Softer rail pads were supposed to solve the vibration problem. When trains pass over tracks, vibration transmits through the rail, through the fastener, into the sleeper, into the ground, and into nearby buildings. Softer pads between rail and sleeper absorb more of this vibration. The fix seemed straightforward.
Softer rail pads reduce ground-borne vibration but increase airborne noise for passengers and bystanders. The mechanism: soft pads allow a greater length of rail to vibrate freely. The rail becomes a longer resonant antenna for sound. More rail vibrating means more sound radiated directly into the air, even though less energy reaches the ground structure.
Conversely, stiffer pads constrain the vibrating length of rail — less airborne noise — but transmit more energy into the ground. You can protect the buildings or protect the ears, but not both.
At Sound Transit, a rough grinding pattern created a persistent 63 Hz resonance spike that lasted years because the light-rail vehicles' axle loads were insufficient to smooth it out. The very trains that created the noise were too light to fix it.
There is no single pad stiffness that optimizes both airborne and ground-borne noise. The energy has to go somewhere. Isolating one path doesn't eliminate the vibration — it redirects it to whichever path you weren't measuring. The engineering tradeoff is not between noise and quiet but between which kind of noise you're willing to accept and who has to hear it.